740795

research-article2017
JFM0010.1177/1098612X17740795Journal of Feline Medicine and SurgeryZeugswetter et al

Original Article

Journal of Feline Medicine and Surgery

Radial vs coccygeal artery Doppler 2018, Vol. 20(10) 968­–972

© The Author(s) 2017
Article reuse guidelines:
blood pressure measurement in sagepub.com/journals-permissions
DOI: 10.1177/1098612X17740795
https://doi.org/10.1177/1098612X17740795

conscious cats journals.sagepub.com/home/jfm

This paper was handled and processed
by the European Editorial Office (ISFM)
for publication in JFMS

Florian K Zeugswetter1, Alexander Tichy2

and Katharina Weber1

Abstract
Objectives  The aims of the study were to compare the results of systolic blood pressure (SBP) measurements at
the radial and coccygeal artery by Doppler ultrasonography and to assess the impact of gender, age, body weight,
body condition score (BCS) and stress score on disagreements.
Methods  Eighty cats were included in this prospective, block randomised crossover trial. Blood pressure
measurements were performed at home, during hospitalisation or in a local animal shelter by a single trained
observer according to the latest ISFM consensus guidelines. Statistical analysis included Spearman’s correlation
and ANOVA.
Results  Measurement or site order had no effect on blood pressure measurements (P = 0.157, P = 0.965).
Correlation of radial and coccygeal SBP was moderate (rsp = 0.519, P <0.001) and measurements at the tail were
by mean 18.7 ± 37 mmHg higher. The differences between the measuring sites were neither correlated with age
(rsp = 0.044, P = 0.710) nor with body weight (rsp = 0.122, P = 0.337). A significant effect of BCS (P = 0.016),
but not gender (P = 0.246) or stress score (P = 0.424) was identified. Of the 25 cats with a coccygeal SBP ⩾170
mmHg, 14 (56%) were categorised as mildly hypertensive or normotensive when measurements were performed
at the forelimb.
Conclusions and relevance  Choice of the measuring site, coccygeal vs radial artery, has a significant impact on
Doppler ultrasonographic SBP results. In many cats, these discrepancies are huge and would certainly affect
treatment decisions. Possible explanations include cone-shaped tails, especially in obese cats, leading to distal
gapping or slippage of cylindrical cuffs and peripheral pulse pressure amplification. Further studies are urgently
needed to clarify whether the discordances are artefacts or have a pathological background.

Accepted: 9 October 2017

Introduction
Doppler-shift sphygmomanometry is well established in
scientific studies on feline hypertension and is likely the
most common indirect systolic blood pressure (SBP)
measurement procedure used in small animal practice.
The technique, although requiring some user skill, is
easy to learn,1 and the equipment relatively inexpensive.
The currently recommended sites for the positioning of
the inflatable cuff are the forelimbs and the tail.2,3 Studies
in anaesthetised cats comparing measurements at the
at the tail are preferred by the majority of cats, resulting
in fewer failures and shorter expenditure of time,6 little
is known about the precision (reproducibility of the
measurements) and accuracy of this measurement site.
Binns et  al7 found good correlation (r = 0.91), good
1ClinicalDepartment for Small Animals and Horses, University
of Veterinary Medicine Vienna, Vienna, Austria
2Platform Bioinformatics and Biostatistics, University of Veterinary

forelimb with direct radiotelemtry (gold standard)
showed good correlation (r = 0.928),4 but a small nega-
tive bias of 11.5–14 mmHg.4,5 With a negative bias of
−22.4 ± 17.1 mmHg and a correlation of 0.855, accuracy
(proximity of the measurements to the true value) was
clearly lower in conscious cats.4 Although measurements
Medicine Vienna, Vienna, Austria
Corresponding author:
Florian K Zeugswetter DrMedVet, PrivDoz, Clinical Department
for Small Animals and Horses, University of Veterinary Medicine
Vienna, Veterinaerplatz 1, Vienna, Austria
Email: florian.zeugswetter@vetmeduni.ac.at
Zeugswetter et al 969
accuracy (mean error −4.7 mmHg), but low precision
(SD ± 22.87 mmHg) when comparing tail measurements
with the gold standard in 11 anaethetised cats. In 28% of
the measurements, differences exceeded 20%.
The pivotal objectives of this study were to compare
radial and coccygeal measurements and their impact
on treatment decisions using current guidelines.3
Additionally, the impact of age, body weight, gender,
anxiety score and body condition score (BCS) on SBP dif-
ferences between the two sites was investigated. During
the performance of this study, Whittemore et  al8 pub-
lished a similar clinical trial including 66 conscious cats.
The authors found only moderate correlation (r = 0.45)
between the results obtained at the different measure-
ment sites, and SBP at the tail was a mean of 19 mmHg
(15%) higher. Worryingly, on the basis of coccygeal artery
results, 28 (42%) of the cats would have been categorised
as hypertensive (>150 mmHg), whereas radial measure-
ments suggested normotension in 20 of these.
Materials and methods
The prospective, crossover study was performed in April
2017 and privately owned cats, hospitalised patients of a
university clinic, as well as cats from a local animal shelter
were included. A stress score from 1 (fully relaxed) to 7
(terrified) as proposed by Kassler and Turner9 was applied.
Cats with a stress score ⩾6, kittens below an age of 6
months, as well as sedated, anaesthetised and acutely
injured cats were excluded.
Measurements were performed as recommended by
the ISFM consensus guidelines3 and the order of meas-
urement was block-randomised. The same Doppler
ultrasonic flow detector (Model 811-B, Eickemeyer,
Parks Medicals Electronics), sphygmomanometer and
neonate disposable blood pressure cuffs (Philips, size
one: 3.1–5.7 cm; size two: 4.3–8.0 cm; size three: 5.8–10.9
cm) were used in all cats. All measurements were per-
formed by a single trained observer, assisted by the cat
owner (privately owned cats) or a nurse (hospitalised or
shelter cats), in a quiet room wearing headphones and
after an acclimatisation period of at least 5 mins. Cats
were allowed to settle into a position of their choosing
and only exceptionally, in case of continuous attempts to
escape, forced into right lateral recumbency. The latter
were assigned a stress score of 5.9 If possible, examina-
tions were performed in sternal recumbency on the lap
of the assistant. Cuff size was selected so that the cuff
width was 30–40% of the circumference of the tail or
forelimb. The hair below the cuff was not removed, and
all measurements were performed within 10 mins. SBP
was calculated from the mean of five consecutive meas-
urements. In contrast to the ISFM guidelines, but based
on the study of Jepson et  al,1 the first of five measure-
ments was not discarded if the variation was ⩽20% of
the second measurement. BCS was determined by using
a five-point scale according to Lund et  al.10 Cats were
grouped according to their mean SBP results, and the
impact on treatment decisions (group 0: <90 mmHg;
group 1: ⩾90 to <150 mmHg; group 2: ⩾150 to 159
mmHg; group 3: ⩾160 to 169 mmHg; group 4: ⩾170
mmHg).3
The study was discussed and approved by the insti-
tutional ethics and animal welfare committee in accord-
ance with Good Scientific Practice guidelines and
national legislation (ETK-11/04/2017).
Statistical analysis was performed with the labora-
tory software package IBM SPSS version 24. Data were
tested for normality of distribution by Shapiro-Wilk
test, and data are given as mean ± SD. To account for
possible effects of measurement order (first vs second
set of readings) or site order (coccygeal/radial vs radial/
coccygeal), analysis with ANOVA for repeated meas-
ures was applied. BCS, gender, and stress score were
included as covariates. For statistical analysis, cats with
a stress score of 1 and 2, as well as 3 and 4 were pooled.
Spearman rank correlation coefficients were calculated
to assess the relationship between the various variables.
For inferential statistics, the level of significance was set
at P <0.05.
Results
Ninety-four cats were enrolled in the study, of which 14
cats were excluded due to intolerance of measurements
(both sites: n = 7, forelimb: n = 5, tail: n = 2). The study
population finally consisted of 80 cats, 68 domestic
shorthairs (85%), four Persians, three British Shorthairs,
two Norwegian Forest Cats, two Maine Coons and one
Bengal; 45 males (six intact) and 35 females (eight intact).
The cats ranged in age from 7 months to 17 years (80.8 ±
52 months) and weighed 2–7.78 kg (4.1 ± 1.3 kg). Forty-
eight cats (60%) were considered to have an optimal BCS
(BCS 3/5), whereas 21 (26%) were under- (BCS 1–2/5),
and 11 (14%) overweight (BCS 4–5/5). Of the 80 cats that
were included in the study, 31 (39%) were healthy and 49
(61%) had one or more known disease processes or
conditions.
Blood pressure measurements were performed at
home in 40 (50%), during hospitalisation in 20 (25%), and
in a local animal shelter in 20 (25%) cats. Fifty-nine (74%),
16 (20%) and five (6%) cats were assigned a stress score
of 1–2 (fully or weakly relaxed), 3–4 (weakly or very
tense) and 5 (fearful, stiff, patience and restraint neces-
sary), respectively. No clear preference for either coccy-
geal or radial measurement was observed in 71 (89%) of
the cats. Five cats showed aversive behaviour towards
coccygeal measurements, whereas four cats disliked
measurements at the paws. Cuff sizes one (forelimb, n =
13; tail, n = 23) and two (forelimb, n = 64; tail, n = 57)
were used in most patients, while cuff size three was
needed in only three cats at the forelimb.
970 Journal of Feline Medicine and Surgery 20(10)
Figure 1  Bland–Altman plot of agreement between forelimb
and tail systolic blood pressure (SBP) measured with Doppler
ultrasonography. The straight black line indicates the mean
bias. The dotted black lines indicate the limits of agreement
(± 2 SD). The broken line indicates zero bias
Measurement order (first vs second set of measure-
ments) or site order (coccygeal/radial vs radial/coccy-
geal) had no effect on blood pressure measurement (P =
0.157, P = 0.965). SBP measurements at the radial artery
(141 ± 32, range 50–280 mmHg) were 18.7 ± 37 mmHg (P
<0.001) lower than at the coccygeal artery (160 ± 45,
range 51–280 mmHg), but discordances occurred in both
directions (see Figure 1). The correlation was moderate
(rsp = 0.519, P <0.001). The differences between the meas-
uring sites were neither correlated with age (rsp = 0.044,
P = 0.710) nor with body weight (rsp = 0.122, P = 0.337).
A significant effect of BCS (P = 0.016, ANOVA, see
Figure 2), but not gender (P = 0.246) or anxiety score (P
= 0.424) was identified. The discordances were higher
in overweight (BCS ⩾4) compared with normal (BCS 3,
P = 0.021) or underweight (BCS ⩽2, P = 0.036) cats.
In 48 (60%) cats, the choice of measurement site
would not have impacted the treatment decision. This
was not the case in 32 (40%) cats where tail measure-
ments were higher (24, 30%) or lower (eight, 10%) than
results obtained at the forelimb. Of the 25 cats with an
SBP ⩾170  mmHg (group 4) measured at the tail, four
(16%), six (24%) and eight (32%) cats were assigned to
groups 3 (⩾160 mmHg), 2 (⩾150 mmHg) and 1 (⩾90 to
<150 mmHg) when measurements were performed at
the forelimb, respectively. One cat was categorised as
hypotensive based on coccygeal (51 mmHg) and radial
(50 mmHg) SBP measurements. Follow-up blood pres-
sure measurements were performed in some of the 20
hospitalised cats, but unfortunately the site of measure-
ment was not recorded. Cardiological examinations
were performed in two of the four cats with severe
hypertension measured at the tail (⩾170 mmHg), and
Figure 2  Box and whisker plots for systolic blood pressure
(SBP) differences between forelimb and tail measurements,
grouped by body condition score (BCS) categories. The lines
within the boxes represent the median and the upper and lower
boundaries of the boxes, the 25th and 75th percentiles. The
whiskers include 25% of the cases, and outliers are depicted
as dots. A significant effect of BCS (P = 0.016, ANOVA)
was identified. The discordances were higher in overweight
(BCS 4 and 5) compared with normal (BCS 3, P = 0.021) or
underweight (BCS 1 and 2, P = 0.036) cats
mild hypertension when measured at the forelimb.
Concentric left ventricular hypertrophy was diagnosed
in one of these patients.
Discussion
The results of this study corroborate the results of Binns
et al7 in anaesthetised and Whittemore et al8 in conscious
cats, which showed that the choice of the measuring site
(forelimb or tail) massively impacts SBP results obtained
with Doppler-shift sphygmomamometry. As in the latter
study, mean SBP measurements at the coccygeal and
radial artery were only moderately correlated (r = 0.519
vs 0.45), and readings at the tail were on average 18.7
mmHg (vs 19 mmHg) higher. Worryingly, of 25 cats with
an SBP ⩾170 mmHg at the tail, 14 (56%) were catego-
rised as mildly hypertensive or normotensive, when
measurements were performed at the forelimb.
Accordingly, the choice of measurement site would have
greatly impacted treatment decisions.
Possible explanations for these discrepancies include,
on the one hand, artefacts caused by body positioning,
stress/discomfort or cone-shaped tails, especially in
overweight animals, and on the other hand true differ-
ences caused by peripheral pulse pressure amplification.
Based on the study of Bodey and Michell11 in dogs,
where significantly lower coccygeal SBP were measured
in a standing position, as compared with lateral recum-
bence, Whittemore et  al8 proposed that perhaps lateral
Zeugswetter et al 971
positioning caused the disagreement. However, as most
of the cats in the current study were allowed to settle in
a sternal position, this hypothesis can be discarded. It is
also unlikely that stress or discomfort was causative, as
no effect of the stress score was observed.
As especially obese animals have cone-shaped tails,
cylindrical cuffs are used. The consequences are distal
gapping, slippage during inflation and, finally, the need
for higher cuff pressures to occlude the arteries. In accord-
ance with this hypothesis, overweight cats had signifi-
cantly higher paw-tail differences than well proportioned
or underweight cats in this study. So called ‘radial’ coni-
cal cuffs are now commercially available for human
patients.12 These special cuffs are the consequence of
studies that demonstrated that up to 15% of obese people
are falsely classified as hypertensive, if traditional cylin-
drical cuffs are used.13,14, Although the average error was
only 2.0 ± 0.4 mmHg in the latter study, differences of up
to 9.7  mmHg were found in individuals with large arms.14
In a recent study including 17 anaesthetised dogs, the use
of specially-fabricated conical cuffs at the limb did not
show any benefit.15 Unfortunately, all of the dogs under
investigation were normotensive, the blood pressure
range was very narrow, the study was not randomised,
and BCS was not given.
Another possible cause for SBP differences between
measurement sites is peripheral pulse pressure (PP)
amplification. It is a clear misconception to believe that
blood pressure is constant, regardless of the location
where it is measured. In short, the pulse waves gener-
ated by the heart travel to the peripheral arteries where
they are reflected. These retrograde waves encounter the
forward pressure waves, which are summated. The con-
sequence is an increase in the whole amplitude of the PP
as it travels distally. The degree of this amplification,
which can reach >30 mmHg as the wave moves from the
aorta to the brachial artery in humans, depends on the
distance to the reflection site, the arterial properties (eg,
elasticity/stiffness, vasomotor tone) and heart rate. As a
consequence, PP amplification has a high inter- and
­intra-individual variability.16–18
In line with this concept, direct measurements at the
median sacral artery or dorsal pedal artery in dogs were
on average 3.08/16.12  mmHg (dorsal recumbence) or
4.67/14.7 mmHg (lateral recumbence) higher than direct
measurements at the superficial palmar arch.19 In
humans, blood pressure differences between the arms of
10 mmHg or more are associated with peripheral vascu-
lar disease with a high specificity. As differences of more
than 15 mmHg are associated with an increased all-cause
and cardiovascular mortality, bilateral brachial measure-
ment has been recommended as a screening test for
peripheral vascular disease.20
Interestingly, BCS influenced the difference between
the measurements at different body sites in this study.
Obesity is clearly associated with cardiovascular disease
in humans21 and, in combination with the metabolic syn-
drome and hypertension, increases PP amplification.18 The
best way to verify different SBPs in the radial and coccy-
geal arteries of affected cats would be to use direct tech-
niques such as radiotelemetry. The main obstacle is the
very small size of the feline coccygeal artery.
Given the discordant SBP results between forelimb
and tail measurements in this and a recent study,8 the
question arises which site to use in future feline studies
and in clinical practice. In our opinion, the recommenda-
tion of Whittemore et al8 to favour the tail is not compel-
ling. Although we agree that measurements at the tail
are very well tolerated, and according to current litera-
ture the accuracy is high (low bias), the precision is low
(high standard deviation). In the only study comparing
Doppler ultrasonographic measurements at the tail with
direct radiotelemetry,7 correlation was 0.91 at the tail and
0.96 at the hindlimb. Twenty-six percent and 21.5% of the
readings had a difference >20 mmHg when measured at
the tail and the hindlimb, respectively. In the same study,
the mean error was −4.7 ± 22.87 mmHg at the tail and
9.4 ± 14.86 mmHg at the hindlimb. In another study
including a radiotelemetry system, measurements at the
forelimbs performed better than at the hindlimbs.4 In
summary, published data do not allow preference of one
or the other site. In our opinion, it is no disadvantage
that forelimb measurements underestimate SBP, as the
diagnostic precision is high4,5,22 and most studies to
determine reference intervals or treatment targets were
performed using this measurement site. In contrast to an
earlier study using high definition oscillometry,6 we
found no clear preference for either coccygeal or radial
measurement.
Conclusions
Given the findings of the study reported here and the
study of Whittemore et al,8 further studies including cats
with a high BCS and the investigation of target organ
damage are urgently needed to clarify whether the con-
spicuous discordances of the SBP between the measure-
ment sites are an artefact or have an anatomical or
pathological background. Until then, the site of measure-
ment should be documented and the results have to be
interpreted with great caution.
Conflict of interest The authors declared no potential
c­onflicts of interest with respect to the research, authorship,
and/or publication of this article.
Funding  The authors received no financial support for the
research, authorship, and/or publication of this article.
References
1 Jepson RE, Hartley V, Mendl M, et  al. A comparison of
CAT Doppler and oscillometric Memoprint machines for
non-invasive blood pressure measurement in conscious
cats. J Feline Med Surg 2005; 7: 147–152.
972 Journal of Feline Medicine and Surgery 20(10)
2 Stepien RL. Feline systemic hypertension: diagnosis and
management. J Feline Med Surg 2011; 13: 35–43.
3 Taylor SS, Sparkes AH, Briscoe K, et  al. ISFM consensus
guidelines on the diagnosis and management of hyper-
tension in cats. J Feline Med Surg 2017; 19: 288–303.
4 Haberman CE, Morgan JD, Kang CW, et al. Evaluation of
Doppler ultrasonic and oscillometric methods in cats.
J Appl Res Vet Med 2004; 2: 279–289.
5 Grandy JL, Dunlop CI, Hodgson DS, et al. Evaluation of the
Doppler ultrasonic method of measuring systolic arterial
blood pressure in cats. Am J Vet Res 1992; 53: 1166–1169.
6 Cannon MJ and Brett J. Comparison of how well conscious
cats tolerate blood pressure measurement from the radial
and coccygeal arteries. J Feline Med Surg 2012; 14: 906–909.
7 Binns SH, Sisson D, Buoscio DA, et  al. Doppler ultraso-
nographic, oscillometric sphygmomanometric and pho-
toplethysmographic techniques for non-invasive blood
pressure measurement in anaesthetised cats. J Vet Int Med
1995; 9: 405–414.
8 Whittemore JC, Nystrom MR and Mywby DI. Effects of
various factors on Doppler ultrasonographic measure-
ments of radial and coccygeal arterial blood pressure in
privately owned, conscious cats. J Am Vet Med Assoc 2017;
250: 763–769.
9 Kassler MR and Turner DC. Stress and adaptation of
cats (Felis silvestris catus) housed singly, in pairs and in
groups in boarding catteries. Anim Welfare 1997; 6: 243–254.
10 Lund EM, Armstrong PJ, Kirk CA, et al. Health status and
population characteristics of dogs and cats examined at
private veterinary practices in the United States. J Am Vet
Med Assoc 1999; 214: 1336–1341.
11 Bodey AR and Michell AR. Epidemiological study of
blood pressure in domestic dogs. J Small Anim Pract 1996;
37: 116–125.
12 Hersh LT, Sesing JC, Luczyk WJ, et al. Validation of a coni-
cal cuff on the forearm for estimating radial artery blood
pressure. Blood Press Monit 2014; 19: 38–45.
13 Maxwell MH, Waks AU, Schroth PC, et al. Error in blood-
pressure measurement due to incorrect cuff size in obese
patients. Lancet 1982: 2: 33–36.
14 Palatini P, Benetti E, Fania C, et al. Rectangular cuffs may
overestimate blood pressure in individuals with large
conical arms. J Hypertens 2012; 30: 530–536.
15 Ramos SJ, Da Cunha AF, Domingues M, et al. Comparison
of blood pressure measurements of anesthetized dogs
obtained noninvasively with a cylindrical blood pressure
cuff and an anatomically modified conical blood pressure
cuff. Am J Vet Res 2015; 77: 59–64.
16 Avolio AP, Van Bortel LM, Boutouyrie P, et  al. Role of
pulse pressure amplification in arterial hypertension.
Experts’ opinion and review of the data. Hypertension
2009; 54: 375–383.
17 Vasan RS. Pathogenesis of elevated peripheral pulse pres-
sure: some reflections and thinking forward. Hypertension
2008; 51: 33–36.
18 Vergnaud AC, Protogerou AD, Li Y, et al. Pulse pressure
amplification, adiposity and metabolic syndrome in sub-
jects under chronic hypertensive therapy: the role of heart
rate. Atherosclerosis 2008; 199: 222–229.
19 Acierno MJ, Dominique ME, Ramos SJ, et al. Comparison of
directly measured arterial blood pressure at various ana-
tomic locations in anesthetized dogs. Am J Vet Res 2014; 76:
266–271.
20 Clark CE, Taylor RS, Shore AC, et al. Association of a dif-
ference in systolic blood pressure between arms with
vascular disease and mortality: a systematic review and
meta-analysis. Lancet 2012; 379: 905–914.
21 Behrenson GS, Srinivasan SR, Bao W, et  al. Association
between multiple cardiovascular risk factors and ath-
erosclerosis in children and young adult. New Engl J Med
1998; 338: 1650–1656.
22 Klevans LR, Hirkaler G and Kovacs JL. Indirect blood pres-
sure determination by Doppler technique in renal hyper-
tensive cats. Am J Physiol 1979; 237: 720–723.